LinuxQuestions.org

LinuxQuestions.org (/questions/)
-   Linux - General (http://www.linuxquestions.org/questions/linux-general-1/)
-   -   Freezes while unattended: where to start looking (http://www.linuxquestions.org/questions/linux-general-1/freezes-while-unattended-where-to-start-looking-4175430713/)

Weapon S 10-05-2012 01:24 PM

Freezes while unattended: where to start looking
 
I'm running Debian stable on quite old hardware.
When I leave my PC in Debian however, it will occasionally freeze. (Alt SysRq not working.) When running Windows XP, it never happens.
So where do I start diagnosing the problem?

NyteOwl 10-05-2012 04:25 PM

Do you sue a screensaver or power management? If so try turning them off first and see if it solves the problem.

H_TeXMeX_H 10-06-2012 05:28 AM

You should check the logs in '/var/log/messages' and syslog.

elmig 10-06-2012 06:32 AM

also the processes that are running:

~# ps aux

Weapon S 10-06-2012 09:56 AM

Thanks for the replies. Screensaver and power management were not enabled. I've checked the root partition with e2fsck -ccfkv; no disk errors.
It might have to do with doing things at 100% CPU. Thunar apparently occasionally crashes when that's true. (I didn't have conky before, but did notice (silent) crashes. Now that I have Conky that seems to be a criterium.)
Quote:

You should check the logs in '/var/log/messages' and syslog.
I had a look. The things that jump out are "quirks" reported: my mobo uses some old vt82c686 chipset or something; and apparently wicd segfaulted once.
Should I remember the times of the crashes, or is there another way to search through those logs? :-/ Searching for "rsyslogd" seems to look up the start of the log (among other things).
BTW hopefully unrelated is that I get a kernel failure or something when I try to mount a cdrom (not auto-detected, blurts out "dunno what iso9660 is" if I try manually). Could it just be the old guy is going to kick the bucket soon? :-/ Maybe I should do a little stress testing in Windows, to exclude that...

H_TeXMeX_H 10-06-2012 10:12 AM

You should look at the log right before the last boot, this was when in crashed. So, just go to the end of the file, then scroll up until you find the beginning of the last boot, when the times suddenly change. When it crashes there likely will be no more output to the logs until the next boot, so look for this time gap.

Weapon S 10-12-2012 12:36 PM

Thanks for your suggestions. It turned out my memory was failing. The condition for hanging system is quite random and more prone to happen when doing disk paging.
Eventually I also encountered trouble in Windows. I should have run memtest earlier... :-X


All times are GMT -5. The time now is 05:23 PM.